Résumé:
Biometric authentication, such as fingerprints and facial recognition, is
increasingly used for securing devices and networks but still poses security risks.
Palmprints offer a promising alternative. In this work, we introduce a new
feature extraction method, Triangular and Orthogonal Local Binary Patterns
(TAO-LBP), based on the LBP descriptor. TAO-LBP improves orientation and
texture coding, enhancing robustness to rotation and orientation changes.
Experiments on IITD and CASIA palmprint datasets show that TAO-LBP
outperforms standard LBP, making it a more reliable method for palmprint
recognition.